Blog
Articles, guides and insights on algorithmic trading, backtesting strategies, quantitative analysis, and data quality.
1–9 of 99 articles
Skip Adapter Work: Cross Asset Data Normalization for Developers
Developer pipeline for cross asset data: harmonize timestamps to ISO 8601 UTC, choose Z score or RobustScaler per series, and cut adapter work with...
What Seven Years of NQ Tick Data Taught Us About Finding Real Edges
A Reddit trader spent months backtesting seven years of NQ tick data and surfaced two candidate strategies. Here's what that process reveals about how much data you actually need, and how to avoid fooling yourself along the way.
Avoid Hour Errors: Three Parts to Normalize Time Zones for Developers
IANA aware dev guide: three part fix to normalize time zones and avoid DST and historic offset errors. Python, JS, Postgres recipes.
Market Data Formats for Developers & Quants: SBE, Parquet, Dollar Bars
A developer and quant primer on market data formats: SBE/FAST feeds, Parquet and CSV storage, tick/volume/dollar sampling, ingestion checks, and...
99% Modeling Quality for Quants: MT5 Backtesting Data IS to WFA to OOS
Trust MT5 backtests: use 99% modeling-quality ticks, set realistic costs, run IS to WFA to OOS validation, and speed final checks with a ready-import...
Avoid Look Ahead Bias: Indices Intraday Data Sourcing for Quants
For quants sourcing intraday index data: focus on point in time membership, documented adjustments, and import ready minute bars to keep backtests honest.
Quants: 20,000 Bars Is Roughly 10 Weeks of Intraday History
Quants: convert platform bar caps into calendar time. 20,000 one minute bars is about 10 weeks. Learn when tick data matters and where to get clean minute...
Avoid 50–100x Slowdowns: MT4 vs MT5 Data Rules for Quants
Engineer checklist for MT4 and MT5 data: match timezones, import M1 or real ticks, use MT5's nine column format, and run a five step audit to stop bad...
Stop Fake Alpha: 5 Checks Quants Need for Hedging Backtesting Data
A quant's vendor-diligence checklist for hedging backtesting data: five field-level checks, four validation tests, and a realistic fills model to avoid...